When AI cuts record review time, who should keep the savings?
A few threads here lately mention big time savings on large-record review, and also worry about what that does to hourly billing. I'm curious how people think the economics shake out.
Under hourly billing, the efficiency gain is awkward. If a review that took 100 hours now takes a fraction of that, the firm bills less for the same output unless rates rise or volume grows. The client sees a faster result and wonders why the invoice didn't shrink more.
There seem to be three options. The firm keeps the savings through fixed fees. The client gets them through lower bills. Or the cost of the tools eats most of it, which some of you have argued.
I work on the commercial side at an AI-native firm, so I see the pricing side of this more than the legal work. On the matters we've scoped as fixed fee, review effort has dropped by 80%, but I'm not sure that tells us much about how a traditional firm should handle it.
For those of you who've seen this up close: do clients actually push for the savings, or do they still just compare hourly rates? And does the answer change for the junior work that used to absorb that review time?